Faster R-CNN Based Defect Detection of Micro-precision Glass Insulated Terminals

DOI オープンアクセス
  • Liu Qunpo
    School of Electrical Engineering and Automation, Henan Polytechnic University
  • Wang Mengke
    School of Electrical Engineering and Automation, Henan Polytechnic University
  • Liu Zonghui
    School of Electrical Engineering and Automation, Henan Polytechnic University
  • Su Bo
    School of Electrical Engineering and Automation, Henan Polytechnic University
  • Naohiko Hanajima
    College of Information and Systems, Muroran Institute of Technology

説明

Micro-precision glass insulated terminals (referred to as glass terminals) are the core components used in precision electronic equipment. As glass terminal, its quality has a huge impact on the performance of precision electronic equipment. Due to limitations in materials and production processes, some of the glass terminals produced have defects such as missing blocks, bubbles, and cracks. At present, it is difficult to ensure product quality and production efficiency with manual inspection methods. However, the defect characteristics of glass terminals are quite different, and it is difficult for traditional defect detection technology to design an ideal feature extractor for detection. Therefore, this paper proposes to use deep learning technology to detect missing blocks. First, preprocess the sample pictures of missing block defects of glass terminals, and then train the deep learning network based on Faster RCNN. According to the test results, the algorithm has an accuracy of over 90% in detecting missing defects in glass terminals.

収録刊行物

詳細情報 詳細情報について

  • CRID
    1390851175702317440
  • DOI
    10.5954/icarob.2021.os6-1
  • ISSN
    21887829
  • 本文言語コード
    en
  • データソース種別
    • JaLC
    • Crossref
    • OpenAIRE
  • 抄録ライセンスフラグ
    使用不可

問題の指摘

ページトップへ